1: %\usepackage{picinpar}
2: \documentstyle[prc,epsfig,aps]{revtex}
3: \def\btt#1{{tt$\backslash$#1}}
4: \newcommand{\gap}{\mathrel{ \rlap{\raise.5ex\hbox{$>$}}
5: {\lower.5ex\hbox{$\sim$}} } }
6: \newcommand{\lap}{\mathrel{ \rlap{\raise.5ex\hbox{$<$}}
7: {\lower.5ex\hbox{$\sim$}} } }
8:
9: \textheight 25 cm
10: \textwidth 17 cm
11:
12: \def\bild#1\over#2{\mathrel{\mathop{\kern5pt #1}\limits_{#2}}}
13: \pagestyle{plain}
14: \newcommand{\fr}{\flushright}
15: \newcommand{\ds}{\displaystyle}
16:
17: \newcommand{\mtca}{\multicolumn{1}{|c}} %{|c} }
18: \newcommand{\mtcb}{\multicolumn{1}{c|} }
19: \newcommand{\be}{\begin{equation}}
20: \newcommand{\ee}{\end{equation}}
21: \newcommand{\ba}{ \begin{eqnarray} }
22: \newcommand{\ea}{ \end{eqnarray} }
23: \newcommand{\bc}{\begin{center}}
24: \newcommand{\ec}{\end{center}}
25: \newcommand{\ca}{$^{14}$C\ }
26: \newcommand{\bea}{$^{14}$Be\ }
27: \newcommand{\bec}{$^{12}$Be\ }
28: \newcommand{\beb}{$^{10}$Be\ }
29: \newcommand{\li}{$^{11}$Li\ }
30: \newcommand{\lic}{$^{9}$Li\ }
31: \newcommand{\cc}{$^{12}$C\ }
32: \newcommand{\cb}{$^{10}$C\ }
33: \newcommand{\plb}[1]{Phys. Lett. B {\bf #1}}
34: \newcommand{\zpb}[1]{Zeit. Phys. Lett. {\bf #1}}
35: \newcommand{\prla}[1]{Phys. Rev. Lett. {\bf #1}}
36: \newcommand{\pr}[1]{Phys. Rev. C {\bf #1}}
37: \newcommand{\praa}[1]{Phys. Rev. A {\bf #1}}
38: \newcommand{\apb}[1]{Ann. of Phys. (N.Y.) {\bf #1} }
39: \newcommand{\npb}[1]{Nucl. Phys. {\bf #1}}
40:
41: \begin{document}
42: \title{\large
43: \bf{Two-body correlations in N=8 and 10 nuclei and effective neutron-neutron
44: interactions in Tamm-Dancoff
45: and two-particle RPA models.}}
46: \author{\normalsize J.C. Pacheco
47: \\
48: {\normalsize Departamento de F\'{\i}sica Aplicada e
49: Instituto de F\'{\i}sica Corpuscular} \\
50: {\normalsize Universidad de Valencia,Burjasot, Valencia, Spain} \\
51: \normalsize N.Vinh Mau \\
52: {
53: \normalsize
54: Institut de Physique Nucl\'eaire,} \\
55: {\normalsize F-91406 , Orsay Cedex, France } \\
56: }
57:
58: %\date{}
59:
60: \maketitle
61: %\noindent PACS numbers : 21.10.Re, 21.60.Jz, 21.90.+f
62:
63: %\noindent IPNO/TH 90-07
64: %\newpage
65: \begin{abstract}
66: We apply a particle-particle RPA model to study the properties of the
67: two-neutron valence wave function in nuclei $^{14}$C, $^{12}$Be, $^{11}$Li and
68: $^{14}$Be. The RPA model takes account of
69: two-body correlations in the cores so that it gives a better description
70: of energies and amplitudes than
71: models which assume a neutron closed shell (or subshell) core.
72: With a Gogny neutron-neutron effective interaction or with the equivalent
73: density dependent delta force we are able to reproduce the two-neutron
74: separation energies in these nuclei and in the corresponding cores,
75: except for $^{9}$Li. These calculations suggest the same 2s-1p$_{1/2}$
76: shells inversion in $^{12}$Be-$^{13}$Be than in $^{11}$Be.
77: \end{abstract}
78:
79: \section{Introduction}
80: In an earlier work \cite{pvm}, we calculated the properties of the two-neutron
81: valence pair in the nuclei $^{11}$Li, $^{12}$Be, $^{14}$C using a Tamm-Dancoff
82: model and assuming the core to have a closed 1p$_{3/2}$ neutron shell.
83: Important ingredients in the calculations are the assumed pairing
84: interaction and the role of the p$_{1/2}$-s$_{1/2}$ shells
85: inversion that is visible in the spectrum of $^{11}$Be.
86:
87:
88: While the model gave a coherent and reasonable description of
89: these nuclei
90: it fails to describe $^{6}$He (taking now an alpha-particle core)
91: giving a two-neutron
92: separation energy of several MeV instead of the experimental value of 0.97
93: MeV \cite{wa}. This has already been found by Esbensen et al. \cite{hb}.
94: When one compares $^{6}$He
95: described as an alpha-particle + two neutrons to \ca for example, one sees
96: immediately an important difference between the two systems: for $^{6}$He the
97: core is well described in an Hartree-Fock model as a
98: pure (1s)$^2_{\nu}$(1s)$^2_{\pi}$ configuration \cite{ca}
99: while we have long known \cite{ck} that
100: in \ca the core of \cc is a mixture of states with neutrons in the
101: 1p$_{3/2}$ or 1p$_{1/2}$ states, what is not taken into account in
102: Tamm-Dancoff models.
103:
104: In this work, we will include the core correlations using the two-particle
105: RPA theory and also examine more broadly the dependence on the assumed
106: effective two-neutron force.
107: Zero range residual forces are very simple to use but are quite arbitrary
108: and binding energies are not sufficient to fix uniquely the force.
109: In a recent work Garrido et al. \cite{gs} have searched for a zero range
110: density dependent force equivalent to the finite range
111: Gogny interaction \cite{gb,jf} (see also \cite{be}).
112: The authors fit the
113: force in order to reproduce the gap calculated in nuclear matter with
114: the Gogny
115: force which has good pairing properties in finite nuclei. A fit to the whole
116: domain of k$_F$ values determines unambigously the parameters of the
117: force and tells what is the cut-off on neutron energy to be used. This last
118: information is very important since a zero range interaction has no natural
119: cut-off. The density independent part of the force
120: reproduces the low energy properties of a free neutron-neutron system so
121: that their force is equivalent to a realistic effective interaction in
122: two-neutron and infinite systems. We will use this force in our problem and
123: will discuss the results compared to the zero range pairing forces used in
124: Tamm-Dancoff models.
125:
126: In section II we briefly report on the results obtained in
127: a pairing or Tamm-Dancoff model with
128: three effective neutron-neutron interactions. In section III we recall
129: the properties and equations of the particle-particle RPA model. The
130: results of this model are presented and discussed in section IV for
131: $^{14}$C-$^{10}$C, $^{12}$Be-$^{8}$Be and $^{11}$Li-$^{7}$Li and in section V
132: for $^{14}$Be-$^{10}$Be with a discussion on the
133: $^{13}$Be and $^{11}$Be spectra. At the end, section VI is devoted to our
134: conclusions.
135:
136:
137: \section{Hamiltonian and effective interactions}
138: We first make a pairing, or Tamm-Dancoff, approximation to describe core + two
139: neutron systems. Assuming an inert and closed
140: sub-shell core for neutrons we
141: diagonalise the two-neutron hamiltonian in a two-neutron subspace built on
142: non occupied neutron states in the core:
143: \ba
144: H_{2n} &=&\frac{{\bf p}_1^2}{2m}+\frac{{\bf
145: p}_2^2}{2m}+V_{nc}(1)+V_{nc}(2)+V_{nn}(1,2)+{\ds \frac{({\bf p}_1+{\bf
146: p}_2)^2}{2 A_c m}}\\
147: &=&h_{nc}(1)+h_{nc}(2)+V_{nn}(1,2)+{\ds \frac{{\bf p}_1.{\bf p}_2}{A_c
148: m}}
149: \ea
150: where the one-neutron hamiltonian is:
151: \ba
152: h_{nc}(i)&=&{\ds \frac{{\bf p}_i^2}{2 \mu}}+V_{nc}(i)\\
153: V_{nc}(r)&=&-V_{NZ}\left (f(r)-0.44r_0^2({\bf {l.s}})\frac{1}{r}
154: \frac{df(r)}{dr}\right)+
155: 16a^2 \alpha_n\;\left(\frac{df(r)}{dr}\right)^2 \\
156: f(r)&=&\left(1+exp(\frac{r-R_0}{a})\right)^{-1}\\
157: V_{NZ}&=&U_0-U_{\tau} \frac{N-Z}{A_c}
158: \ea
159: $\mu$ is the reduced mass equal to ${\ds\frac{A_c\;m}{A_c+1}}$; $A_c$, N, Z are
160: respectively the
161: mass, neutron and proton numbers in the core; $R_0=r_0A_c^{1/3}$ with
162: $r_0$=1.27 fm, $a$=0.75 fm; the strengths $U_{0}$ and $U_{\tau}$ are
163: taken the same as
164: in our previous papers \cite{pvm,sl}. The last term of potential, eq.(4),
165: simulates
166: particle-phonon couplings contribution to the one-body potential \cite{nvm}.
167: The strengths $\alpha_n$ are fitted in each nucleus to reproduce the
168: experimental 1p$_{1/2}$, 2s, 1d$_{5/2}$ single neutron
169: energies for \ca and \bec and in order to get the measured
170: two-neutron separation energy in \li and \bea for 1p$_{1/2}$ and 2s states
171: which are not experimentally well known . Their numerical values can be found
172: in refs.[1] and [10]. For higher neutron states the
173: particle-phonon couplings are weak and we take $\alpha_n$=0. We use a
174: discretisation of the continuum states with a radial box of 20 fm and
175: orthonormalise the wave
176: functions by the Schmidt method. In our previous
177: papers the two-body term ${\bf p}_1.{\bf p}_2$ was neglected but as the
178: approximation was the same for all nuclei: $^{14}$C, $^{12}$Be, $^{11}$Li and
179: $^{14}$Be and the effective neutron-neutron interaction fitted to describe
180: the properties of $^{14}$C and $^{12}$Be, then used in the other
181: nuclei, the effect of such an approximation was minimised.
182: We have checked that adding this
183: term changes slightly the strength of the effective interaction (by few \%)
184: but gives the same agreements and predictions than previously.
185:
186: We used in refs.[1,10] a zero range density dependent neutron-neutron
187: interaction given by:
188: \be
189: V_{nn}(1,2)=-V_0\;\left (1-x\;[{\rho_c( { \frac{{\bf r_1+r_2}}{2}})}/
190: {\rho_0}]^p\right )\delta({\bf r_1-r_2})
191: \ee
192: where $\rho_c$ is the core density and $\rho_0$=0.16fm$^{-3}$.
193: The above fit gives now $V_0$=880 MeV.fm$^{3}$,
194: x=0.93 and p=1.2 close to what is used in the
195: literature \cite{be,tf}. With the same force we
196: have calculated $^{6}$He as an alpha-particle plus two neutrons. Taking
197: $V_{nc}$ as the neutron-alpha particle potential fitted
198: by Satchler et al. \cite{sao} to low energy l=1 phase shifts we get a
199: two-neutron separation energy of about 4 MeV in $^{6}$He while it is
200: 0.97 MeV experimentally \cite{wa}. Similar strong binding was
201: found by Esbensen et al. \cite{hb}
202: who
203: modify the n-n interaction in order to get the experimental value. However
204: this is not satisfactory for us because we always require that our effective
205: interaction leads to an overall agreement for all systems since the density
206: dependent term should take account of the dependence of the effective
207: interaction upon the nucleus.
208: We have looked for different possible sources of
209: such unexpected disagreement: choice of neutron-alpha particle interaction,
210: discretisation of the continuum which will be discussed in a forthcoming paper
211: and effect of the cut-off on neutron energy.
212: However none of these effects is responsible for such a bad result.
213:
214: The force of eq.(7) has three parameters which are not independently
215: determined from a fit of energy spectra.
216: In a recent paper Garrido et al. \cite{gs} have fitted the
217: three
218: parameters of $V_{nn}$ in order to get the same gap, $\Delta(k_F)$, in
219: nuclear matter than obtained with a Gogny finite range
220: effective interaction \cite{gb,jf}. They have shown that to get
221: agreement over all the
222: domain of $k_F$ they have to take p=0.47, x=0.45 with a
223: cut-off, $\epsilon_c^0$, on the neutron energy of 50-60 MeV.
224: The cut-off energy determines
225: the strength $V_0$ if one assumes that the density independent part of the
226: interaction should reproduce the properties of a free neutron-neutron
227: system. It gives the following relation between $V_0$, $\epsilon^0_{c}$
228: and the neutron-neutron scattering length $a_{nn}$ \cite{hb} :
229: \ba
230: V_0&=&2{\pi}^2\;\frac{\hbar^2}{m}\;\frac{1}{k_c^{(0)}-{\ds \frac
231: {\pi}{2a_{nn}}}}\\
232: {k_c^{(0)}}^2&=& \frac{2m\epsilon_c^0}{\hbar^2}
233: \ea
234: $a_{nn}$ is experimentally -18.5 fm then very large. If we replace it by
235: -$\infty$ and take $\epsilon^0_c$=60 MeV this relation gives
236: $V_0$ = 480 MeV.fm$^{3}$ as used in ref.[6].
237:
238: For free neutrons, the neutron energy
239: is only kinetic energy while in finite nuclei the neutrons are in the potential
240: $V_{nc}$.
241: Therefore when using eqs.(8-9) the cut-off energy has to be calculated from the
242: bottom of the single-particle well and $\epsilon^0_c$ replaced by :
243: \be
244: \epsilon^0_c=\epsilon_c+V_{NZ}
245: \ee
246: where $V_{NZ}$ is the depth of our one-body potential, eq.(6),
247: and $\epsilon_c$ the cut-off energy for a neutron described by the
248: hamiltonian of eq.(3).
249: In our calculations we take neutron states up to an
250: energy $\epsilon_c \simeq$
251: 10 MeV for all nuclei. $V_{NZ}$ depends on proton and
252: neutron numbers in the core and varies between about 40 MeV for \li
253: and 50 MeV for \ca, giving an
254: equivalent cut-off in free two-neutron system or nuclear matter of 50-60 MeV
255: as required by the fit of ref.[6]. Note that taking $\epsilon_c$ the
256: same for all nuclei implies that $V_0$ will be slightly different
257: for $^{11}$Li, \bec and $^{14}$C.
258:
259: This new zero range effective interaction has very different parameters
260: compared to the usual ones. We have made the same calculations as previously
261: with this force for our systems $^{14}$C, \bec and $^{6}$He. \li results
262: are not reported in the table but with neutron energies close to
263: experimental values it is not bound. The results
264: are reported in Table I where they are compared to the measured two-neutron
265: separation energy, $S_{2n}$ \cite{wa}.
266: We see two interesting facts: the calculated
267: $S_{2n}$ in $^{6}$He is now 1.02 MeV close to the experimental value but it is
268: much too low in the other nuclei.
269: Since this zero range force is constructed
270: to reproduce the Gogny force in nuclear matter and in a free two-neutron
271: system we thought that this equivalence could fail in our finite
272: nuclei which are quite far from both systems.
273:
274: Then we have made the same Tamm-Dancoff calculation with the genuine
275: Gogny force, D1 or D1S \cite{gb,jf}. The Gogny forces are the sum of central
276: density dependent and spin-orbit terms. For two neutrons coupled to 0$^+$
277: the spin-orbit term gives very small contribution to pairing matrix elements
278: and can be neglected \cite{gg}. Therefore we have made our calculations,
279: keeping the central term only. The results are very close
280: for the two forces and we present them for the force D1S only.
281: The results with the Gogny force are presented in Table I.
282:
283:
284: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
285: \begin{center}
286: \begin{tabular}{|c|c|c|c|} \hline
287: Force &$^{14}$C &$^{12}$Be & $^{6}$He\\ \hline
288: Gogny &11.8 & 2.23& 0.94 \\
289: $\delta$ force & 11.7 & 1.95 & 1.02\\
290: exp. & 13.12 & 3.67 & 0.97\\
291: \hline
292: \end{tabular}
293: \vskip 6mm
294: TABLE I. Two-neutron separation energies in MeV calculated in Tamm-Dancoff
295: model with the Gogny DS1 force and a zaro range force of eq.(7) with
296: $V_{0}$=480~MeV.fm$^{3}$, p=0.47 and x=0.45
297: \end{center}
298: \vskip 4mm
299: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
300:
301:
302:
303:
304:
305: We see that they
306: are the same as obtained with its zero range substitute what tells us that:
307:
308: 1- the equivalence between the two forces, with zero or finite range, shown
309: for a free system and nuclear matter holds in all considered finite nuclei.
310: We give the results for $V_0$=480 MeV.fm$^3$, the same for all nuclei,
311: but as already mentionned $V_0$ should be larger in \bec than in \ca
312: following relations (6), (8) and (10) what will
313: improve the agreement between the two series of results.
314:
315: 2- the range of the effective force is not responsible for the unability of
316: the nuclear model to describe simultaneously $^{6}$He and the p-shell nuclei.
317:
318: The agreement for
319: $^{6}$He on one side, the disagreement for the other nuclei on the
320: other side seem
321: at first sight a contradiction of the model. However it can
322: be understood when we go back to the first assumption of the Tamm-Dancoff
323: model that the core is inert with neutrons filling the lowest shells.
324: For $^{6}$He the core,
325: an
326: alpha-particle, is strongly bound and very well described in an Hartree-Fock
327: model with neutrons and protons filling the 1s shell \cite{ca}.
328: It means that the
329: assumption of a closed shell core (implicitly assumed when we put our extra-
330: neutrons on the unoccupied shells with probability one) is valid in this case.
331: However for the other nuclei we know that the cores are not good closed shell
332: nuclei. It is known for long in the case of \cc from the work of Cohen-Kurath
333: \cite{ck}. For \bec shell model calculations also show a large deviation from
334: a pure state as it is required by experiments \cite{so,ba,ss,na}. It is also
335: obvious in our calculations for \bec \cite{pvm} or in three-body Fadeev model
336: calculations of Thompson et al. \cite{tn} where \bec is described as a core of
337: \beb plus two neutrons and found as a large mixture of (2s)$^2$,
338: (1p$_{1/2}$)$^2$ and (1d$_{5/2}$)$^2$ two-neutron components while it is
339: considered as a pure (2s)$^2$ or (1p$_{1/2}$)$^2$ state when one studies
340: $^{14}$Be.
341:
342: A model which takes account of two-body correlations in the ground
343: state of the core is
344: the particle-particle RPA model more generally called pair vibration model
345: \cite{sa,rp,bvm}. We
346: briefly recall what this model is for two neutrons outside a core which, in
347: an independent particle model, would have closed shells (or subshells) for
348: neutrons.
349:
350: \section{Particle-particle RPA model}
351:
352: Particle-particle RPA relies on the expansion of the two-particle
353: (two-hole) Green's function in terms of ladder diagrams with upward and
354: backward going diagrams as shown in Fig.1. Note that a summation over
355: upward going diagrams only leads to the Tamm-Dancoff approximation of the
356: previous section. From the related approximate integral equation satisfied
357: by this RPA two-particle Green's function, a system
358: of equations is derived which
359: describes simultaneously the $A_c$+2 and $A_c$-2 systems where $A_c$
360: characterises the core which in an independent particle model would be
361: described as a closed shell (or sub-shell) nucleus.
362:
363:
364: %$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
365: \begin{figure}
366: \begin{center}
367: \begin{tabular}{c}
368: \epsfysize=5cm
369: \epsfxsize=10cm
370: \epsfbox{fig1.eps}
371: \end{tabular}
372: \vskip 1mm
373: FIG.1: RPA diagrammatic expansion of the two-particle Green's function.
374: \end{center}
375: \end{figure}
376: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
377:
378:
379: We here apply this method to our problem of two neutrons outside a core with
380: $N_c$ neutrons assuming that the protons are not disturbed when one adds or
381: subtracts two neutrons.
382:
383: Let's define a=(a$_1$,a$_2$), b=(b$_1$,b$_2$)$...$ two-neutron configurations
384: with the neutrons in states a$_1$,a$_2$$...$ unoccupied in the Hartree Fock
385: core ground state, $\alpha$,$\beta$$...$ two-neutron configurations with the
386: neutrons in occupied states $\alpha_1$,$\alpha_2$$...$ and two-neutron
387: amplitudes or spectroscopic factors as:
388: \ba
389: X_{\alpha}(J,M)&=&<N_c+2|A^+_{\alpha}(J,M)|N_c,0>\\
390: X_a(J,M)&=&<N_c+2|A^+_a(J,M)|N_c,0>
391: \ea
392: where the pair operators are given by:
393: \ba
394: A^+_a(J,M)&=&\sum_{m_{a_1},m_{a_2}}(j_{a_1},j_{a_2},m_{a_1},m_{a_2}|J,M)
395: a^+_{a_1}a^+_{a_2}\;\;\;\;\,\,with \,\,a_1\leq a_2\\
396: A^+_{\alpha}(J,M)&=&\sum_{m_{\alpha_1},m_{\alpha_2}}(j_{\alpha_1},
397: j_{\alpha_2},m_{\alpha_1},m_{\alpha_2}|J,M)
398: a^+_{\alpha_1}a^+_{\alpha_2}\;\;\;\;\,\,with\,\,\alpha_1\leq\alpha_2
399: \ea
400: $a^+_i$ is the creation operator of a neutron in state $j_i,m_i$,
401: $|N_c+2>$ and $|N_c,0>$ are respectively
402: the RPA wave functions of the $N_c+2$ nucleus
403: in excited or ground state and of the core in its ground state.
404: We see immediately that the amplitudes $X_{\alpha}$ are non zero
405: only if the core
406: ground state has 2p-2h,4p-4h,\ldots
407: components. Assuming that all $X_\alpha$ are zero, one gets back to the
408: Tamm-Dancoff approximation.
409:
410: In the same way we define two-neutron hole amplitudes for the $N_c-2$ nuclei
411: from $A_a$ and $A_{\alpha}$, the annihilation operators for a pair which
412: are hermitian
413: conjugates of $A^+_a$ and $A^+_{\alpha}$ defined by eqs.(11-12). They are:
414: \ba
415: Y_a(J,M)=<N_c-2|A_a(JM)|N_c,0>\\
416: Y_{\alpha}(J,M)=<N_c-2|A_\alpha(J,M)|N_c,0>
417: \ea
418:
419: For the $N_c$-2 nuclei $Y_a$ are non zero only if the core ground state
420: is not a pure HF state but has 2p-2h, 4p-4h\ldots components. Then
421: these anomalous components $X_{\alpha}$ and $Y_a$ give a measure of the
422: two-body correlations in the cores. For a given spin and parity
423: the RPA amplitudes $X$ and $Y$ satisfy the same system of equations:
424: \ba
425: (E-\epsilon_a)x_a-\sum _b<a|V_{nn}|b>x_b-\sum_{\beta}
426: <a|V_{nn}|\beta>x_{\beta}&=&0\\
427: (E-\epsilon_{\alpha})x_{\alpha}+\sum_b<\alpha|V_{nn}|b>x_b+\sum_{\beta}
428: <\alpha|V_{nn}|\beta>x_{\beta}&=&0
429: \ea
430: where $x$ are the amplitudes X or Y as explained below.
431: The two-body matrix elements are antisymmetrised and $\epsilon_a$,
432: $\epsilon_{\alpha}$
433: the sum of unperturbed energies of the two neutrons in configurations a,
434: $\alpha$ respectively.
435: The one-neutron states are described by the one-body effective potential of
436: eq.(4) what means that we expand our two-body Green's function of Fig.1 in
437: terms of one-body propagators which include particle-phonon
438: couplings as given in Fig.2.
439:
440: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
441: \begin{figure}[h]
442: \begin{center}
443: \begin{tabular}{c}
444: \epsfysize=5cm
445: \epsfxsize=10cm
446: \epsfbox{fig2.eps}
447: \end{tabular}
448: \vskip 1mm
449:
450: FIGURE 2. One-particle Green's function with inclusion of particle-phonon
451: coupling diagrams.
452: \end{center}
453: \end{figure}
454: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
455:
456:
457: If we take N configurations a,b.. and M configurations $\alpha$, $\beta$,..
458: the equations (17,18) have N+M solutions. N solutions correspond to
459: the $N_c+2$ nucleus with eigenvalues $E_{nJ}$ and amplitudes $X^{(n)}$
460: where n labels the state such that:
461: \ba
462: E_{nJ}&=&E_{nJ}(N_c+2)-E_0(N_c)\\
463: X_a^{(n)}(J)&=&x_a^{(n)}\;\;\;,\;X_\alpha^{(n)}(J)=x_\alpha^{(n)}
464: \ea
465: the amplitudes satisfy the following orthonormalisation relation:
466: \be
467: \sum_a X_a^{(n)}\;X_a^{(n')}-\sum_\alpha X_{\alpha}^{(n)}\;X_\alpha^{(n')}=
468: \delta_{nn'}
469: \ee
470: The M other solutions (labelled by the index m) are eigenstates
471: of the $N_c-2$ nucleus with
472: eigenvalues $E_{mJ}$ and amplitudes $Y$ given by:
473: \ba
474: {\cal E}_{mJ}&=&(E_{mJ}(N_c-2)-E_0(N_c))=-E_{mJ}\\
475: Y_a^{(m)}(J)&=&x_a^{(m)}\;\;\;,\;Y_\alpha^{(m)}(J)=x_\alpha^{(m)}
476: \ea
477: with the orthonormalisation condition:
478: \be
479: \sum_a Y_a^{(m)}\;Y_a^{(m')}-\sum_\alpha Y_{\alpha}^{(m)}\;Y_\alpha^{(m')}
480: =-\delta_{mm'}
481: \ee
482: We see from eqs.(19) and (22) that the lowest energies
483: $E_{00}$ and ${\cal E}_{00}$
484: give the two-neutron separation energy in the $N_c+2$ and $N_c$ nuclei
485: respectively with the relations:
486: \ba
487: S_{2n}(N_c+2)&=&-E_{00}\\
488: S_{2n}(N_c)&=&{\cal E}_{00}
489: \ea
490:
491: Then, contrary to particle-hole RPA, all solutions of eqs.(17,18)
492: correspond to
493: physical states. The separation between N and M states follows from the
494: relative importance of $x_a$ and $x_\alpha$ amplitudes: for the $N_c+2$
495: nucleus the amplitudes $x_a$ are larger than $x_\alpha$ and inversely for
496: $N_c-2$ nucleus.
497:
498: This model is now applied to the p-shell nuclei. As noticed already
499: $^{6}$He described as an $\alpha $ particle plus two neutrons is very likely
500: well reproduced by a Tamm-Dancoff model since the core is a good Hartree-Fock
501: system. On the other hand, the $N_c$-2 nucleus (di-proton) is not a bound
502: system and the RPA model is not reliable in this case so that we discuss the
503: p-shell nuclei only.
504:
505: \section{Results for N=8 nuclei: $^{14}$C, \bec and \li}
506:
507: We make the calculation for the two effective interactions, the Gogny finite
508: range interaction and the zero range interaction of eq.(7)
509: with the parameters p and x
510: fitted by Garrido et al. and a strength $V_0$ fitted to give the same RPA
511: energies as the Gogny force. Afterwards we shall compare
512: with the strength given by eqs.(8-10).
513: The two-neutron subspace (a,b\ldots) is the same as in section II with a
514: normal ordering of 1p$_{1/2}$-2s shells in C-isotopes but an inversion
515: of these two shells in Be and Li.
516: For the occupied neutron states ($\alpha$, $\beta$\ldots )
517: we take the two shells 1s$_{1/2}$ and 1p$_{3/2}$
518: . The 1p$_{3/2}$ neutron energy is taken as the known one-neutron
519: separation energy for the corresponding core (\cc or \beb or \lic)\cite{wa} .
520: For the 1s shell we have no
521: experimental information and we take it as a parameter by changing the depth
522: of our Saxon-Woods potential. This very deep state
523: has very little effect on the $N_c$+2 states but a non negligible effect on
524: the energy of the $N_c$-2
525: ground states. Then we take the 1s energy in order to get close agreement,
526: if possible, with the experimental two-neutron separation energy
527: in the core nucleus. This procedure gives $\epsilon$(1s)=-32, -32 and -28
528: for $^{12}$C, $^{10}$Be and $^{9}$Li respectively.
529:
530:
531:
532: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
533:
534: \begin{center}
535: \begin{tabular}{|c|cc|cc|cc|}
536: \hline
537: & $^{14}$C & $^{12}$C & $^{12}$Be &$^{10}$Be & $^{11}$Li & $^{9}$Li \\ \hline
538: Gogny &12.9 & 32. & 3.69& 8.2 & 0.37 & 3.8 \\
539: $\delta$ force & 12.9 & 32.5 & 3.6 & 8.52 & 0.34 & 3.76\\
540: exp. & 13.12 & 31.8 & 3.67 & 8.48 & 0.34 $\pm $0.05 &6.1\\
541: \hline
542: \end{tabular}
543: \vskip 6mm
544: TABLE II. Same as Table I for RPA model and $V_0$=500, 510 and 560~MeV.fm$^{3}$
545: for $^{14}$C-$^{12}$C, $^{12}$Be-$^{10}$Be and $^{11}$Li-$^{9}$Li respectively
546: \end{center}
547: \vskip 4mm
548: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
549:
550:
551:
552:
553:
554:
555: In Table II are presented $S_{2n}$, the two-neutron separation energies
556: in $N_c$+2 ($^{14}$C, $^{12}$Be and $^{11}$Li) and $N_c$ ($^{12}$C, $^{10}$Be
557: and $^{9}$Li) nuclei given by the lowest eigenvalues $E_{00}$ and
558: ${\cal E}_{00}$ respectively, following relations (25) and (26). The results
559: are given for the Gogny force
560: and the equivalent zero range force. In Table III we
561: give the amplitudes $X_a$ and $X_{\alpha}$ for $N_c+2$ nuclei and $Y_a$ and
562: $Y_{\alpha}$ for $N_c-2$ nuclei corresponding to the ground states and
563: obtained with the Gogny interaction. The zero range interaction gives nearly
564: identical eigenvectors.
565:
566: For $^{14}$C-$^{12}$C and $^{12}$Be-$^{10}$Be, $S_{2n}$ is close to the
567: experimental values while we have no free parameters, apart from the
568: 1s-energy which turns out from the fitting procedure to be very close to a
569: typical Hartree-Fock energy. We see that the amplitudes $X_{\alpha}$
570: for \ca and
571: \bec are for $\alpha=(1p_{3/2})^2$ rather large and larger in \bec
572: than in $^{14}$C. This large value means that in the ground state of the cores,
573: $^{12}$C and $^{10}$Be, there are large components of 2nh-2np states with at
574: least two
575: holes in the $1p_{3/2}$ shell. This is in complete agreement with the shell
576: model results for \cc \cite{ck} and for \bec \cite{so,ba,ss,na}.
577: Furthermore we see that, if in \ca
578: the amplitude for the two neutrons in a $1p_{1/2}$ state is very large, in
579: \bec we have comparable amplitudes for $(2s)^2$, $(1p_{1/2})^2$ and
580: $(1d_{5/2})^2$ configurations. This means that, together with the fact that
581: $X_{\alpha}=0.46$, \bec cannot reasonably be considered as a closed shell
582: nucleus as done in many papers on $^{14}$Be.
583:
584:
585: Our RPA equations give simultaneously the amplitudes $Y_{a}$ and
586: $Y_{\alpha}$ of eqs.(15-16) for \cb and $^{8}$Be. Here the $Y_a$ are the
587: anomalous amplitudes reflecting again correlations in the cores. We see
588: consistency with the results for $^{14}$C and $^{12}$Be. The amplitudes $Y_a$
589: are larger in $^{8}$Be than in \cb with a distribution over several two-neutron
590: states revealing a very complicated structure of $^{8}$Be.
591:
592: The $^{11}$Li-$^{7}$Li systems present more ambiguity than the
593: previous ones. Now we know from break-up experiments that the lowest
594: neutron resonance in $^{10}$Li is an s-state at 0.1-0.2
595: MeV \cite{pf} as required by several calculations \cite{pvm,tz,bh}
596: and that the next state is a p$_{1/2}$ resonance at
597: 0.54$\pm$ 0.06 MeV \cite{li}.
598:
599:
600:
601: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
602: \begin{center}
603: \begin{tabular}{|c|cccc|cc|}\hline
604: &\multicolumn{4}{c|}{$X_a(Y_a)$} &\multicolumn{2}{c|}{$X_\alpha
605: (Y_{\alpha})$}\\ \hline
606: & (2s)$^2$ & (1p$_{1/2})^2$ & (1p$_{1/2}$,2p$_{1/2}$) & (1d$_{5/2})^2$ &
607: (1p$_{3/2})^2$ & (1s)$^2$ \\ \hline
608: $^{14}$C &-0.12&0.96&-0.05&-0.28&0.19&-0.09\\
609: $^{10}$C &-0.05&0.15&-0.02&-0.14&0.98&-0.32\\ \hline
610: $^{12}$Be & -0.48&0.76&-0.29&-0.44&0.60&-0.09\\
611: $^{8}$Be & -0.16&0.34&-0.17&-0.30&1.18&-0.12\\ \hline
612: $^{11}$Li & 0.66&-0.56&0.48&0.04&-0.45&0.07\\
613: $^{7}$Li & -0.12&0.22&-0.27&0.02&1.12&-0.11\\
614: \hline
615: \end{tabular}
616: \vskip 6mm
617: TABLE III. The most important RPA amplitudes ($X_a,X_\alpha $ for $N_c+2$
618: nuclei, $Y_a ,Y_\alpha$ for $N_c-2$ nuclei)
619: \end{center}
620: \vskip 4mm
621: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
622:
623:
624:
625:
626:
627:
628: The results, presented in Tables II and III, have been obtained with
629: $\epsilon(1s)=-28$ MeV, $\epsilon(2s)=0.19$MeV and $\epsilon(p_{1/2})=0.6$
630: MeV close to measurements for the two last ones and reasonable for
631: the first one. We first see that $S_{2n}$ in \li is
632: in good agreement with the experimental value, 0.34$\pm$0.05 MeV \cite{lia}.
633: The anomalous
634: amplitude $X_{\alpha}$ for $\alpha$=(1p$_{3/2}$)$^2$ is as large as it was
635: in \bec revealing strong deviation from closed shell in $^{9}$Li. We also see
636: that the amplitudes for (1p$_{1/2}$)$^2$
637: and (2s)$^2$ configurations are
638: similar, as required by the analysis of break-up experiments \cite{br}.
639: However
640: the $S_{2n}$ value for $^{9}$Li is very far from the measured value, 3.85 instead
641: of 6.1 MeV. This disagreement, found only in Li, cannot be improved by
642: a reasonable change of $\epsilon$(1s) which is the only parameter of the
643: calculation. It could come from the single last proton which
644: is on the same shell as the active neutrons in $^{7}$Li and \lic but we did
645: not find any simple way to evaluate such an effect.
646:
647: In the case of Li isotopes, there is an ambiguity coming from higher neutron
648: states. Experimentally one has not seen the d$_{5/2}$ resonance, therefore we
649: have calculated it with the bare Woods-Saxon potential even though
650: we know that in
651: $^{11}$Be one has to add a surface term which lowers the d$_{5/2}$ energy.
652: However the effect of such a term should not affect the results since, as
653: seen in Table III, the amplitude for the (1d$_{5/2})^2$ configuration is
654: very small. Also, because $^{10}$Li
655: is unbound, all neutron states are in the continuum with the result
656: that continuum non-resonant states play a more important role in the
657: determination of the \li energy. This effect is
658: particularly important for the discretised 2p$_{1/2}$ state which is
659: taken as an
660: eigenstate of the bare Woods-Saxon potential. Because the modification of
661: the potential by particle-vibration couplings is very large for the
662: 1p$_{1/2}$ resonance, this 2p$_{1/2}$ state comes at an energy close to
663: the 1p$_{1/2}$, giving a large overlap between the two states. We know that
664: particle-vibration
665: couplings are more important for states close to the Fermi
666: surface and can be neglected for higher states in usual nuclei. However in
667: $^{10}$Li these couplings are very strong and very likely still efficient
668: for the 2p$_{1/2}$ state. Then as a test we have made the calculation with
669: the same surface contribution to the
670: one body potential for
671: 1p$_{1/2}$ and 2p$_{1/2}$ states, taking the bare potential for the other
672: p-states which anyway have much higher energies and therefore have a weaker
673: effect on the ground state energy and wave function. The amplitude $X_a$ for
674: a=(1p$_{1/2}$,2p$_{1/2}$) is weaker, $S_{2n}$ slightly smaller. To recover
675: the same value one has to take
676: $\epsilon$(1p$_{1/2}$)=0.53 MeV, then at the experimental energy, but
677: these differences are
678: not significant.
679:
680: Going back to Table II we see that
681: the strengths $V_0$ of the zero range interaction fitted to recover the same
682: energy than the Gogny interaction are 500, 510 and 560 MeV.fm$^{3}$ for
683: $^{14}$C, $^{12}$Be and $^{11}$Li respectively. They follow closely
684: the dependence on $V_{NZ}$ given by
685: eqs.(8-10) with a value for \ca very close to 480 MeV.fm$^{3}$ used
686: in ref.[6]. We again find the same equivalence
687: between the two forces,with zero or finite range, as
688: derived for free neutrons and nuclear matter. It seems from our study that
689: the need of a much stronger zero range force to get experimental binding
690: energies in simple pairing models, as reminded in Section II, is in
691: fact a way to take implicitly account of two-body correlations in the core
692: which are neglected in the model.
693:
694:
695: \section{Results for \bea (N=10).}
696:
697:
698: The problem of \bea is not yet completely clarified, both experimentally and
699: theoretically. Experimentally one knows that $^{13}$Be has a
700: d$_{5/2}$ resonance at 2.01 MeV
701: above the n+$^{12}$Be threshold \cite{ost}, a lower resonance at about
702: 0.8 MeV seen in the reaction $^{14}$C($^{11}$B,$^{12}$N)$^{13}$Be \cite{pos}
703: which has no spin or parity assignment and, from a recent experiment \cite{th}
704: using fragmentation of
705: $^{18}$O and detecting neutrons in coincidence with $^{12}$Be, a 1/2$^+$
706: resonance below 0.2 MeV which should be the ground state of this
707: unbound nucleus. Theoretical models describing \bea as two neutrons
708: outside a core of $^{12}$Be, where the neutrons fill the 1s, 1p$_{3/2}$ and
709: 1p$_{1/2}$ shells, have either to lower the d$_{5/2}$ resonance or to assume
710: a bound 2s state therefore to bind $^{13}$Be to get a correct binding energy
711: in \bea \cite{tn}. In a first paper \cite{sl}
712: based on a two-neutron Tamm-Dancoff model and with a zero range effective
713: interaction fitted on \ca and \bec we found that an inversion of the
714: 1p$_{1/2}$-2s shells leads to the correct binding
715: energy in \bea without modifying the known d$_{5/2}$ energy nor assuming a
716: bound
717: $^{13}$Be. This calculation, as others, assumes a closed shell nucleus of
718: \bec where the neutrons fill the 1s-1p$_{3/2}$-2s shells,
719: while we have seen above that the ground state of \bec has amplitudes
720: $X_a$
721: over several two-neutron configurations and, because of large $X_\alpha$, has
722: very likely components on more complicated
723: configurations. The
724: particle-particle RPA applied to \bea takes this into account
725: and will provide a description
726: of the core consistent with the RPA amplitudes of Table III.
727:
728: We have made the RPA calculation assuming a normal ordering of shells:
729: 1s, 1p$_{3/2}$, 1p$_{1/2}$, 2s,\ldots with a Hartree Fock state where the p-shell
730: is filled for neutrons. With $\epsilon$(1p$_{1/2}$)=-3.15 MeV, given by the
731: neutron separation energy in $^{12}$Be \cite{wa}, a 1d$_{5/2}$ state at 2 MeV,
732: the
733: experimental value, and a 2s state at 7 keV, very close to the threshold,
734: we get S$_{2n}$=0.7 MeV much too low compared to the experimental value of
735: 1.34$\pm$0.11 MeV \cite{wa}. For \bec we get S$_{2n}$=3.26 MeV, also
736: too small. Moreover the amplitudes $X_{\alpha}$ are small indicating
737: weak correlations in \bec in disagreement with the results of the previous
738: section. Therefore we find again that we are not able to
739: describe satisfactorly \bea when we assume a normal ordering of
740: shells in $^{12}$Be.
741:
742:
743: We now assume an inversion of the two shells 2s and 1p$_{1/2}$ as in $^{11}$Be.
744: The configurations $\alpha$ are built on 1s,1p$_{3/2}$ and 2s states while
745: the configurations a include neutrons in 1p$_{1/2}$ state.
746: The results are summarised in Table IV where we give S$_{2n}$ for
747: \bea and the RPA amplitudes for $^{14}$Be and $^{10}$Be. They are
748: obtained for a Gogny effective interaction with the d$_{5/2}$ resonance at
749: the experimental energy of 2 MeV, a 1p$_{1/2}$ resonance at 0.68 MeV, an
750: occupied 2s shell with an energy of -3.15 MeV given by
751: the experimental neutron
752: separation energy in \bec and a 1p$_{3/2}$ state at -5.6 MeV.
753: For the $\alpha$ configurations we have taken the
754: (1p$_{3/2}$)$^2$, (1s,2s) and (2s)$^2$ states.
755: We have checked that adding the (1s)$^2$
756: states does not change our results.
757:
758: We get 3.65 MeV for the two-neutron separation energy in $^{12}$Be. By
759: comparing with the result of Table II we see that the values
760: of S$_{2n}$ in \bec found in both calculations
761: are very close showing the coherence of the model. Indeed we get 3.65 MeV when
762: \bec is considered as the core for the $^{14}$Be-$^{10}$Be systems while it is
763: 3.69 MeV when it is described as two neutrons outside a core of $^{10}$Be.
764:
765: The p$_{3/2}$ energy cannot be deduced directly from experiments.
766: However $^{11}$Be has a 3/2$^-$ state at 3.9 Mev excitation energy \cite{aj},
767: then at 7 MeV when referred to the \bec core, with a very small width and has
768: certainly a large component of one neutron-hole in $^{12}$Be
769: mixed with
770: 2h-1p components which have higher energies. Therefore a hole energy of
771: 5.6 MeV is not unrealistic.
772:
773: According to our discussion for $^{11}$Li we have done the calculation
774: using the same surface potential for 1p$_{1/2}$ and 2p$_{1/2}$ states. We
775: have then to lower the 1p$_{1/2}$ energy to 0.56 MeV if we leave all other states
776: to be the same , what does not modify qualitatively our conclusions.
777:
778:
779:
780:
781: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
782:
783: \begin{center}
784: \begin{tabular}{|c|c|ccc|cc|}\hline
785: & S$_{2n}$(MeV) & (1p$_{1/2})^2$ & (1p$_{1/2}$,2p$_{1/2}$)& (1d$_{5/2})^2$ &
786: (1p$_{3/2})^2$ & (2s)$^2$ \\ \hline
787: $^{14}$Be & 1.30 & -0.46 & 0.62 & 0.74 & -0.64 & 0.46 \\
788: $^{10}$Be & - & -0.20 & 0.31 &0.46 & -0.73 & 1.\\
789: \hline
790: \end{tabular}
791: \vskip 6mm
792: TABLE IV. RPA energy in MeV for \bea and main amplitudes for \bea and \beb
793: obtained with the D1S Gogny force.
794: \end{center}
795:
796: \vskip 4mm
797:
798: %&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
799:
800:
801:
802: The amplitudes given in Table IV show again strong correlations in $^{12}$Be.
803: Indeed the
804: anomalous amplitudes in \bea are 0.64 and 0.46 for (p$_{3/2})^2$ and (2s)$^2$
805: respectively what means that in \bec components on configurations with two
806: holes on these shells are important. Note that the components with two holes
807: in the 2s-shell are qualitatively consistent with the
808: amplitudes $X_a$ of Table III for a=(1p$_{1/2}$)$^2$ and (1d$_{5/2})^2$.
809: However RPA gives only overlap of wave functions and a strict and direct
810: comparison between amplitudes derived in the two calculations is not
811: possible. To make a direct comparison one has to calculate wave functions
812: what can be made using a quasi-boson approximation as was done in the past for
813: particle-hole RPA correlations.
814:
815: Our fitted value $\epsilon(1p_{1/2})$=0.68 MeV is consistent with an unbound
816: 1/2$^-$ state in $^{13}$Be which would be at 0.68 MeV above the \bec+n
817: threshold, close to the experimental resonance at 0.8 MeV. However the
818: recent experiment using fragmentation of $^{18}$O \cite{th} shows
819: a low energy ($\leq$0.2 MeV) s-wave strength in $^{13}$Be what, in
820: an independent
821: particle model, would mean that the lowest unoccupied shell in \bec is an s
822: shell and would reject the possibility of inversion. In our model however
823: $^{13}$Be is described as a neutron added to the correlated core of
824: $^{12}$Be which is a mixture of many different states, in particular
825: it has large components on configurations with
826: a closed 1p$_{3/2}$ shell plus two neutrons on the 2s or on the 1p$_{1/2}$
827: shell. Therefore these two components will give in $^{13}$Be two different
828: states, a 1/2$^-$ state built on the first one with the last
829: neutron on the p$_{1/2}$ shell (the 2s-shell is filled)
830: and a 1/2$^+$ state built
831: on the second one with the last neutron on the empty 2s
832: shell. In a weak coupling model, because of the known inversion in
833: n+$^{10}$Be system, one can show that the 1/2$^+$ state is
834: lower than the 1/2$^-$ state by about 0.32 MeV and at about 0.3 MeV
835: above the $^{12}$Be+n threshold what is in qualitative agreement
836: with the experimental state seen recently. Consequently
837: a 1/2$^+$ ground state in $^{13}$Be is not in contradiction
838: with an inversion of the 2s-1p$_{1/2}$ shells.
839:
840: There are several other arguments in favor of this inversion in
841: $^{12}$Be-$^{13}$Be. The first one relies on the recent measurement of
842: the B(E2) for the 2$^+$ state at 2.1 MeV in \bec which is found to be
843: the same as in $^{10}$Be for the 2$^+$ state at 3.4 MeV \cite{bei}. Because
844: the inversion in $^{11}$Be is related to the large value of the B(E2) in
845: $^{10}$Be \cite{nvm,nun} there is no reason why the effect
846: should be smaller in $^{13}$Be. Moreover the phonon has a smaller energy in
847: \bec than in $^{10}$Be what is expected to give an
848: enhancement of the coupling \cite{nvm}.
849: Note that this large B(E2) was predicted in ref.[10].
850: Further arguments are found by looking at the $^{11}$Be spectrum. Indeed if
851: $^{13}$Be can be described as a neutron added to a core of $^{12}$Be, $^{11}$Be
852: can be described as a hole in the same core of $^{12}$Be. Therefore the 1/2$^+$
853: ground state of $^{11}$Be is expected to correspond mainly to a hole in
854: the last occupied shell in the Hartree-Fock ground state of \bec what implies
855: that this shell should be an s shell, not a p$_{1/2}$.
856: The first excited states can be obtained as two holes coupled
857: to 0$^+$ plus a neutron on the p$_{1/2}$ or the d$_{5/2}$ shell
858: therefore will be a 1/2$^-$ and a higher 5/2$^+$ states respectively,
859: in agreement with
860: the experimental spectrum. If we assume that the residual interaction
861: between the two holes and the particle are similar if the particle is on
862: a p$_{1/2}$ or d$_{5/2}$ shell we find a difference between the excitation
863: energies of the two states of 1.32 MeV while experimentally it is 1.45 MeV.
864: One sees that the inversion is able to give a coherent description of
865: $^{11}$Be, $^{13}$Be and \bea.
866:
867:
868:
869: Because any model relies on approximations one should be aware of the
870: difficulty to draw a precise scheme. For this reason it is desirable to
871: compare results of different models. Our results without inversion of shells
872: agree with those of Thompson and Zukhov \cite{tn}. We now
873: make a comparison with the work of
874: Descouvemont \cite{deb,de}. He has calculated in the generator
875: coordinate model (GCM) $^{13}$Be
876: and \bea as \bec+n and \bec+n+n systems respectively. The core of \bec is
877: described by a filled 1p-shell for the neutrons while the two protons of the
878: 1p-shell can couple to different states, ground and excited 0$^+$, 1$^+$ and
879: 2$^+$ states. Qualitatively it is equivalent to our Tamm-Dancoff approach
880: where contribution of core excited states are put in our one-body neutron
881: potential and where the neutrons of the core are assumed to fill the 1p-shell.
882: The GCM calculations lead to a slightly bound 1/2$^+$ ground state
883: for $^{13}$Be and a
884: \bea bound by 1.1 MeV. In our Tamm-Dancoff approach with the zero range
885: force fitted on $^{14}$C and $^{12}$Be, a 2s neutron state at -90 keV and a
886: filled neutron 1p-shell (therefore without inversion) we get for \bea a
887: binding energy of -0.93 MeV what is close to the GCM result. The three
888: models, three-body Faddeev, GCM or simple pairing, lead to similar results
889: but are not able to give rise to good agreement for
890: $^{13}$Be and \bea when in \bec the neutrons are assumed to fill the
891: 1p$_{3/2}$-1p$_{1/2}$ shells.
892:
893: A different work \cite{ma} using a density dependent relativistic mean field
894: model calculates
895: $^{12}$Be-$^{14}$Be assuming closed 1p and 1p-2s neutron shells
896: respectively. It gives too large one-
897: neutron and two-neutron separation energies in both systems. This result,
898: together with the assumption of closed shells in both $^{12}$Be and $^{14}$Be,
899: is not in agreement with other calculations.
900:
901: \section{Conclusions}
902:
903: We have first shown that a two-neutron Tamm-Dancoff model with a zero range
904: density dependent neutron-neutron interaction fitted on \ca and \bec gives
905: simultaneously good results for \li and \bea but fails to describe $^{6}$He.
906: The zero range force necessary to get agreement in C-Be-Li nuclei has very
907: different
908: parameters compared to the parameters fitted by Garrido et al. to reproduce
909: the gap calculated in nuclear matter with the finite range Gogny effective
910: interaction. The same Tamm-Dancoff calculation with these two forces shows
911: that they are still equivalent in our finite nuclei and gives a
912: good binding energy in $^{6}$He but too weak binding in N=8 nuclei.
913: This result is well understood in terms of two-body correlations in the
914: cores. Indeed we know that the alpha-particle is well described in
915: Hartree-Fock model. Then two-body correlations in $^{4}$He are unefficient
916: while we know for long from shell model calculations that the cores of \cc
917: and \bec and very likely $^{9}$Li are not pure closed shell nuclei as assumed in
918: a pairing model. It is also obvious for \bec in our calculation: when it is
919: described as a core of \beb plus two neutrons its wave function is a mixture
920: of (2s)$^2$, (1p$_{1/2}$)$^2$ and (1d$_{5/2}$)$^2$ two-neutron states while
921: in the study of \bea it is considered as a pure (2s)$^2$ or
922: (1p$_{1/2}$)$^2$ state what yields inconsistency of the model.
923:
924: The particle-particle RPA model is well adapted to take into account such
925: correlations and indeed the model applied to
926: $^{14}$C-$^{11}$Li-$^{12}$Be-$^{14}$Be gives
927: now with those realistic forces very good agreement with experimental
928: binding energies. It gives also the
929: two-neutron separation energy in the cores. For $^{12}$C-$^{10}$Be and \bec
930: (the latter being considered
931: as a core in the calculation of $^{14}$Be) the agreement with measurements is
932: also
933: very good. However it is too small in $^{7}$Li very likely due to the single
934: proton in the p$_{3/2}$ shell. The model gives also two-neutron
935: and two-neutron hole amplitudes in the wave functions (spectroscopic factors)
936: which are related to the amount of two-body correlations introduced in the
937: cores. This effect is always large but larger in $^{10}$Be-$^{12}$Be than in
938: $^{12}$C, in qualitative agreement with shell model calculations.
939:
940: To get a good two-neutron separation energy in \bea and to get a consistent
941: description of \bec when it is considered as the core of \bea or described
942: as \beb + two neutrons, one has to assume an inversion of 2s-1p$_{1/2}$
943: shells in $^{12}$Be-$^{13}$Be as it is in $^{10}$Li and $^{11}$Be.
944: This inversion is also suggested by
945: a recent measurement \cite{bei} of the transition 2$^+$(2.1 MeV)
946: $\rightarrow$ 0$^+$(gs) in $^{12}$Be. The B(E2) is found to be the same as for the
947: transition 2$^+$(3.3 MeV) $\rightarrow$ 0$^+$(gs) in \beb,
948: suggesting the same effect of particle-phonon couplings in the two systems,
949: $^{11}$Be-$^{13}$Be,
950: therefore the same shells inversion.
951: Furthermore strong particle-particle RPA correlations are known to modify the
952: one-neutron mass operator \cite{vm,bd} therefore to give further
953: corrections to the
954: one-neutron single energies. They may
955: enhance the inversion process studied in refs.[11,34] for $^{11}$Be,
956: even though when
957: adding the two contributions due to couplings with phonons and pair
958: vibrations one has to substract the second order term in order to eliminate
959: double counting, so that only very mixed RPA states will contribute. From the
960: calculated amplitudes one may expect this contribution to be non
961: negligible in $^{11}$Be and $^{13}$Be and even larger in $^{13}$Be than in
962: $^{11}$Be.
963:
964: One of us (NVM) is very grateful to P. Schuck for very fruitful discussions
965: which have initiated an important part of this work.
966: We also are indebted to G.F. Bertsch for
967: many comments and suggestions during the preparation of the manuscript.
968: The second version of the manuscript has been carefully read by F.
969: Bortignon, D. Kadi-Hanifi and P. Schuck. We thank them for accepting this
970: tedious job and for their helpful remarks.
971:
972:
973:
974: \begin{thebibliography}{50}
975: \bibitem{pvm} N. Vinh Mau and J.C. Pacheco, \npb {A607}, 163 (1996).
976: \bibitem{wa} G. Audi and A.H. Wapstra, \npb {A565}, 66 (1993).
977: \bibitem{hb} H. Esbensen, G.F. Bertsch and K. Hencken, \pr {56}, 3054 (1997)
978: \bibitem{ca} X. Campi, J. Martorell and D W.L. Sprung, \plb {41}, 443
979: (1972).
980: \bibitem{ck} S. Cohen and D. Kurath, \npb {73}, 1 (1965).
981: \bibitem{gs} E. Garrido, P. Sarriguren, E. Moya de Guerra and P. Shuck, \pr
982: {60}, 064312 (1999).
983: \bibitem{gb} J. Decharge and D. Gogny, \pr {21}, 1568 (1980).
984: \bibitem{jf} J.F. Berger, M. Girod and D. Gogny, Comp. Phys. Comm. {\bf 63},
985: 365 (1991).
986: \bibitem{be} G.F. Bertsch and H. Esbensen, \apb {209}, 327 (1991).
987: \bibitem{sl} M. Labiche, F.M. Marques, O. Sorlin and N. Vinh Mau, \pr {60},
988: 027303 (1999).
989: \bibitem{nvm} N. Vinh Mau, \npb {A592}, 33 (1995)
990: \bibitem{tf} J. Terasaki, H. Flocard, P.H. Heenen and P. Bonche, \npb {A621},
991: 706 (1997).
992: \bibitem{sao} R. Satchler, L.W. Owen, A.J. Elwyn, G.L. Morgan and R.L. Walter,
993: \npb {A112}, 1 (1968).
994: \bibitem{gg} M. Girod and B. Grammaticos, \pr {27}, 2317 (1983).
995: \bibitem{so} T. Suzuki and T. Otsuka, \pr {56}, 847 (1997)
996: \bibitem{ba} B.A. Brown, International School of Heavy Ion Physics: Exotic
997: Nuclei, edited by R.A. Broglia and P.G. Hansen (World Scientific,
998: Singapore, 1998) p.1
999: \bibitem{ss}H. Sagawa, T. Suzuki, H. Iwasaki and M. Ishihara, \pr
1000: {63}, 034310(2001).
1001: \bibitem{na} A. Navin et al., \prla {85}, 266 (2000).
1002: \bibitem{tn} I.J. Thompson and M.V. Zukhov, \pr {53}, 708 (1996).
1003: \bibitem{sa} N. Fukuda, F. Iwamoto and K. Sawada, \praa {135},
1004: 932 (1964).
1005: \bibitem{rp} G. Ripka and R. Padjen, \npb {A132}, 489 (1969).
1006: \bibitem{bvm} A. Bouyssy and N. Vinh Mau, \plb {35}, 269 (1971); \npb
1007: {A224}, 331 (1974).
1008: \bibitem{pf} S. Pita, These, Universite Paris 6 (2000)
1009: \bibitem{tz} I.J. Thompson and M.V. Zukhov, \pr {49}, 1904 (1994).
1010: \bibitem{bh} G.F. Bertsch, K. Hencken and H. Esbensen, \pr {57}, 1366 (1998)
1011: \bibitem{li} R.M. Young et al., \pr {49}, 279 (1994).
1012: \bibitem{lia} T. Kobayashi, \npb {A538}, 343c (1992).
1013: \bibitem{br} M. Zinzer et al., \prla {75}, 1719 (1995).
1014: \bibitem{ost} A.N. Ostrowski et al., \zpb {A343}, 489 (1992).
1015: \bibitem{pos} A.V. Belozyorov et al., \npb {A636}, 419 (1998).
1016: \bibitem{th} M. Thoennessen, S. Yokoyama and P.G. Hansen, \pr {63},
1017: 014308 (2000).
1018: \bibitem{aj} F. Ajzenberg-Selove, \npb {A506}, 1 (1990).
1019: \bibitem{bei} H. Iwasaki et al., \plb {481}, 77 (2000).
1020: \bibitem{nun}F.M. Nunes et al., \npb {A609}, 43 (1996).
1021: \bibitem{deb} P. Descouvemont, \plb {331}, 271 (1994).
1022: \bibitem{de} P. Descouvemont, \pr {52}, 704 (1995).
1023: \bibitem{ma} Zhongzhou Ren, Gongou Xu, Baoqiu Chen, Zhongyu Ma and W. Mittig
1024: \plb {351}, 11 (1995).
1025: \bibitem{vm} N. Vinh Mau, Theory of nuclear structure: Trieste Lectures
1026: 1969,(IAEA Vienna) (1970) p.931
1027: \bibitem{bd} C. Barbieri and W.H. Dickhoff, \pr {63}, 034313 (2001).
1028: \end{thebibliography}
1029:
1030: \end{document}
1031: